跳到主要內容

簡易檢索 / 詳目顯示

回結果列表
研究生： 楊舒雯
Shu-Wen Yang
論文名稱： 基於深度學習之雷射熱裂應力異常預測
Deep Learning-Based Prediction of Stress Anomalies in Laser Thermal Cracking
指導教授： 林錦德
Chin-Te Lin
口試委員:
學位類別： 碩士
Master
系所名稱： 工學院 - 機械工程學系
Department of Mechanical Engineering
論文出版年： 2025
畢業學年度： 114
語文別： 英文
論文頁數： 111
中文關鍵詞： 雷射熱裂加工 、光彈性影像 、應力分析 、深度學習 、ConvLSTM 、異常預測 、結構相似度指標 、玻璃
外文關鍵詞： Laser Thermal Cracking, Photoelastic Imaging, Stress Analysis, Deep Learning, ConvLSTM, Anomaly Prediction, SSIM, Glass
相關次數： 點閱：10  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 雷射熱裂（Laser Thermal Cracking, LTC）為一種應用於脆性材料精密加工的非接觸式切割技術。即使在相同加工參數條件下，裂縫的傳播行為仍可能因材料內部應力場與熱擴散之不穩定性而產生顯著差異，導致裂縫停滯、熱損傷，甚至切割失敗等異常情形。傳統方法多著重於加工參數的調控，缺乏即時掌握材料內部應力變化的能力，限制了對異常狀況的早期識別與精準控制。
    本研究導入卷積長短期記憶網路（ConvLSTM）進行時間序列影像預測。模型訓練階段分別採用平均平方誤差（Mean Squared Error, MSE）與結構相似度指標（Structural Similarity Index, SSIM）作為損失函數與性能評估指標，以同時評估影像亮度與結構還原的準確性。
    實驗結果顯示，當採用 MSE 作為損失函數時，所建構之 ConvLSTM 模型能有效預測裂縫應力場之時序變化，其測試樣本平均 SSIM 為 0.842、MSE 為 0.0281，整體預測結果穩定，且能有效保留原始影像的結構特徵。進一步分析指出，在約 77% 的測試樣本中，模型預測結果通常可較實際影像提前一影格出現應力變化。本研究進一步整合既有之「影像結構特徵指標系統」，包含亮度、輪廓面積、長寬比與影像重心等多項特徵，達到於異常發生前0.8秒成功預測之結果。

    Laser Thermal Cracking (LTC) is a non-contact cutting technique for precision machining of brittle materials. Even under the same processing parameters, the crack propagation behavior may still be significantly different due to the instability of the internal stress field and heat dissipation of the material, resulting in abnormal situations such as crack stagnation, thermal damage, and even cutting failure. Traditional methods focus on the regulation of processing parameters and lack the ability to immediately grasp the changes in the internal stress of the material, which limits the early identification and precise control of abnormal conditions.
    In this study, a convolutional short-term and long-term memory network (ConvLSTM) is used to predict time series images. In the model training stage, Mean Squared Error (MSE) and Structural Similarity Index (SSIM) are used as the loss function and performance evaluation index to evaluate the accuracy of image brightness and structural restoration, respectively.
    The experimental results show demonstrate that the ConvLSTM model can effectively predict the temporal changes of in the crack stress field when MSE is used employed as the loss function, and the yielding an average SSIM and MSE are of 0.842 and 0.0281, respectively. The overall prediction results are stable, and the structural features of the original images are effectively preserved. It is further analyzed that in about 77% of the test samples, the model predictions usually show the stress changes one frame earlier than the actual images. This study further integrates the existing classification framework, which includes brightness, contour area, aspect ratio, and image center of gravity, to achieve a successful prediction result of 0.8 seconds before the occurrence of the anomaly.

    聲明 i 中文摘要 ii Abstract iii 致謝 iv Table of Contents v List of Figure viii List of Table x Chapter 1 Introduction 1 1-1 Background 1 1-2 Motivation 2 1-3 Summary of Proposed Framework 3 1-4 Objectives 4 1-5 Thesis Structure 4 Chapter 2 Related Principles and Technologies 6 2-1 Principles of Laser Thermal Cracking 6 2-1-1 LTC Stress Field and Fracture Mode 6 2-1-2 Crack Propagation Mechanism 8 2-2 Photoelastic Observation 9 2-2-1 Birefringence and Stress-Optic Law 9 2-2-2 Isoclinic Observation and Image Characteristics 11 2-2-3 Theoretical Basis for Stress Analysis Using Photoelasticity 12 2-3 Machine Learning 13 2-3-1 Fundamentals of Machine Learning 14 2-3-2 Introduction to Deep Learning 15 2-3-3 Convolutional Long Short-Term Memory Networks 17 2-3-4 Performance Evaluation 19 Chapter 3 Methodology 21 3-1 Experimental Setup 21 3-1-1 Overview of Observation System 21 3-1-2 Laser 22 3-1-3 Camera 24 3-1-4 Substrate 25 3-2 Model Architecture 26 3-2-1 Model Requirements 26 3-2-2 ConvLSTM Model Architecture 27 3-2-3 Model Parameters 29 3-3 Experimental Design 30 3-3-1 Four types of Cracks 30 3-3-2 Parameter Selection and Design 33 3-4 Data Preprocessing 35 3-4-1 Region of Interest 36 3-4-2 Spatiotemporal Preprocessing and Illumination Normalization 37 3-4-3 Sample Distribution under Different Scan Speeds 39 3-5 Training and Evaluation 40 3-5-1 Training Strategy 41 3-5-2 Loss Functions 41 3-5-3 Evaluation Approaches 43 Chapter 4 Results and Discussion 45 4-1 Crack Observation and Analysis 45 4-2 Model Evaluation 47 4-2-1 Loss Functions 47 4-2-2 Trained Model Evaluation 49 4-3 Feasibility Analysis of Crack Prediction 52 4-3-1 Evaluation of Visual Performance 52 4-3-2 Prediction via Identification of Crack Behaviors 60 4-4 Anomaly Prediction 61 4-5 Discussion 65 Chapter 5 Conclusion 67 5-1 Contributions 67 5-2 Limitations and Future Work 68 References 72 Appendix 76 A. Photoelastic Feature Indices 76 A.1 Symmetry Axis Detection and Secondary Cropping 76 A.2 Indices Definitions and Computation Logic 79 A.3 Brightness 80 A.4 Contour 82 A.5 Aspect Ratio 84 A.6 Centroid Position 85 B. Experimental Procedure 88 B.1 Experimental Procedure 88 B.2 Image Recording and Storage Settings 89 C. ConvLSTM Code 90

    [1] R. Lumley, "Controlled separation of brittle materials using a laser," Ceramic Bulletin, vol. 48, no. 9, p. 850, 1969.
    [2] 俊. 沖山, "レーザ割断," 精密工学会誌, vol. 60, no. 2, pp. 196-199, 1994, doi: 10.2493/jjspe.60.196.
    [3] J. Powell, CO2 laser cutting, 2 ed. Springer London, 1993.
    [4] 黒部利次, "YAG レーザによる精密割断技術," 精密工学会誌, vol. 65, no. 11, pp. 1556-1559, 1999, doi: 10.2493/jjspe.65.1556.
    [5] W. M. Steen and J. Mazumder, Laser material processing, 4 ed. springer science & business media, 2010.
    [6] K. E. Hazzan, M. Pacella, and T. L. See, "Laser processing of hard and ultra-hard materials for micro-machining and surface engineering applications," Micromachines, vol. 12, no. 8, p. 895, 2021, doi: 10.3390/mi12080895.
    [7] T. Ueda, K. Yamada, K. Oiso, and A. Hosokawa, "Thermal stress cleaving of brittle materials by laser beam," CIRP Annals, vol. 51, no. 1, pp. 149-152, 2002, doi: 10.1016/S0007-8506(07)61487-5.
    [8] H. MORITA, "Crack extension induced by thermal stresses associated with uniform heating in a circle," Journal of Japan Society of Mechanical Engineers, Series A, vol. 56, no. 524, pp. 850-855, 1990.
    [9] G. R. Irwin, "Analysis of stresses and strains near the end of a crack traversing a plate," 1957, doi: 10.1115/1.4011547.
    [10] T.-g. Zhai, A. Wilkinson, and J. Martin, "A crystallographic mechanism for fatigue crack propagation through grain boundaries," Acta materialia, vol. 48, no. 20, pp. 4917-4927, 2000, doi: 10.1016/S1359-6454(00)00214-7.
    [11] A. K. Dubey and V. Yadava, "Laser beam machining—A review," International Journal of Machine Tools and Manufacture, vol. 48, no. 6, pp. 609-628, 2008, doi: 10.1016/j.ijmachtools.2007.10.017.
    [12] K. YAMADA, T. MAEDA, R. TANAKA, and K. SEKIYA, "1210 Photoelastic Observation of Thermal Stress in Laser Cleaving of Hard Brittle Materials," in Proceedings of International Conference on Leading Edge Manufacturing in 21st century: LEM21 2015.8, 2015: The Japan Society of Mechanical Engineers, p. 1210.
    [13] K. Yamada, T. Maeda, T. Iwai, K. Sekiya, and R. Tanaka, "Photoelastic observation of stress distributions in laser cleaving of glass substrates," Precision Engineering, vol. 47, pp. 333-343, 2017, doi: 10.1016/j.precisioneng.2016.09.007.
    [14] 楊舒雯, "Real-time Anomaly Detection in Laser Thermal Cracking by Analysis of Photoelastic Images ", 機械加工システム研究室, 広島大学, 2025.
    [15] J. Orr and J. Finlay, Photoelastic stress analysis (Optical measurement methods in biomechanics). Springer, Boston, MA, 1997, pp. 1-16.
    [16] M. Ayatollahi and M. Nejati, "Experimental evaluation of stress field around the sharp notches using photoelasticity," Materials & Design, vol. 32, no. 2, pp. 561-569, 2011, doi: 10.1016/j.matdes.2010.08.024.
    [17] X. Shi, Z. Chen, H. Wang, D.-Y. Yeung, W.-K. Wong, and W.-c. Woo, "Convolutional LSTM network: A machine learning approach for precipitation nowcasting," in Advances in neural information processing systems, 2015, vol. 28.
    [18] 岡工太郎, "岡工太郎 修士論文," 機械加工システム研究室, 広島大学, 2023.
    [19] H. Aben and C. Guillemet, Photoelasticity of glass, 1 ed. Springer Berlin, Heidelberg, 2012.
    [20] 西田正孝, 材料力学: 光弾性で補説する. 森北出版, 1977.
    [21] A. Schuster, An introduction to the theory of optics, 3 ed. E. Arnold, 1904.
    [22] W. Peters and W. Ranson, "Digital imaging techniques in experimental stress analysis," Optical engineering, vol. 21, no. 3, pp. 427-431, 1982, doi: 10.1117/12.7972925.
    [23] D. Zhang, K. Yamada, Y. Nakajima, R. Tanaka, and K. Sekiya, "Precise Identification of Principal Stress Directions Induced in Laser Cleaving Process," in International Conference on Leading Edge Manufacturing/Materials and Processing, 2020, vol. 83624: American Society of Mechanical Engineers, p. V001T09A005.
    [24] S. Timoshenko, "Goodier. JN, Theory of Elasticity," New. York McGraw—Hil1, vol. 970, no. 4, pp. 279-291, 1970.
    [25] Y. Yamamoto and N. Tokuda, "Determination of stress intensity factors in cracked plates by the finite element method," International Journal for Numerical Methods in Engineering, vol. 6, no. 3, pp. 427-439, 1973.
    [26] J. Clark, A. J. Durelli, and V. J. Parks, "Photoelastic Study of High-Frequency Stress Waves Propagating in Bars and Plates," 1968.
    [27] K. Ramesh and G. Lewis, "Digital photoelasticity: advanced techniques and applications," Appl. Mech. Rev., vol. 55, no. 4, pp. B69-B71, 2002.
    [28] A. Luckow et al., "Artificial intelligence and deep learning applications for automotive manufacturing," in 2018 IEEE International Conference on Big Data (Big Data), 2018: IEEE, pp. 3144-3152, doi: 10.1109/BigData.2018.8622357.
    [29] J. Hirschberg and C. D. Manning, "Advances in natural language processing," Science, vol. 349, no. 6245, pp. 261-266, 2015, doi: 10.1126/science.aaa8685.
    [30] P.-H. Chou, K. Yamada, Y.-R. Hwang, E. Sentoku, R. Tanaka, and K. Sekiya, "Machine Learning for Scanning Path Prediction in Laser Forming-Application of Structured Patterns and CNN," Journal of Laser Micro/Nanoengineering, vol. 19, no. 2, 2024, doi: 10.2961/jlmn.2024.02.3001.
    [31] T. M. Mitchell, Machine learning, 1 ed. (no. 9). McGraw-hill New York, 1997.
    [32] J. Alzubi, A. Nayyar, and A. Kumar, "Machine learning from theory to algorithms: an overview," in Journal of physics: conference series, 2018, vol. 1142: IOP Publishing, p. 012012, doi: 10.1088/1742-6596/1142/1/012012.
    [33] D. E. Rumelhart, G. E. Hinton, and R. J. Williams, "Learning representations by back-propagating errors," nature, vol. 323, no. 6088, pp. 533-536, 1986, doi: 10.1038/323533a0.
    [34] A. Krizhevsky, I. Sutskever, and G. E. Hinton, "Imagenet classification with deep convolutional neural networks," Advances in neural information processing systems, vol. 25, 2012, doi: 10.1145/3065386.
    [35] L. Medsker and L. C. Jain, Recurrent neural networks: design and applications, 1 ed. CRC press, 1999.
    [36] S. Hochreiter and J. Schmidhuber, "Long short-term memory," Neural computation, vol. 9, no. 8, pp. 1735-1780, 1997, doi: 10.1007/978-3-642-24797-2_4.
    [37] S. Ji, W. Xu, M. Yang, and K. Yu, "3D convolutional neural networks for human action recognition," IEEE transactions on pattern analysis and machine intelligence, vol. 35, no. 1, pp. 221-231, 2012, doi: 10.1109/TPAMI.2012.59.
    [38] G. Bertasius, H. Wang, and L. Torresani, "Is space-time attention all you need for video understanding?," in International Conference on Machine Learning, 2021, vol. 2, no. 3, p. 4, doi: doi.org/10.48550/arXiv.2102.05095.
    [39] Z. Liu et al., "Video swin transformer," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 3202-3211.
    [40] A. Vaswani et al., "Attention is all you need," Advances in neural information processing systems, vol. 30, 2017.
    [41] X. Liang, L. Lee, W. Dai, and E. P. Xing, "Dual motion GAN for future-flow embedded video prediction," in proceedings of the IEEE international conference on computer vision, 2017, pp. 1744-1752, doi: doi.org/10.48550/arXiv.1708.00284.
    [42] D. Tran, L. Bourdev, R. Fergus, L. Torresani, and M. Paluri, "Learning spatiotemporal features with 3d convolutional networks," in Proceedings of the IEEE international conference on computer vision, 2015, pp. 4489-4497, doi: doi.org/10.48550/arXiv.1412.0767.
    [43] A. Dosovitskiy et al., "An image is worth 16x16 words: Transformers for image recognition at scale," arXiv preprint arXiv:2010.11929, 2020, doi: doi.org/10.48550/arXiv.2010.11929.
    [44] Y.-C. Hsu, C.-H. Yu, and M. J. Buehler, "Using deep learning to predict fracture patterns in crystalline solids," Matter, vol. 3, no. 1, pp. 197-211, 2020, doi: 10.1016/j.matt.2020.04.019
    [45] Y. Al Najjar, "Comparative analysis of image quality assessment metrics: MSE, PSNR, SSIM and FSIM," International Journal of Science and Research (IJSR), vol. 13, no. 3, pp. 110-114, 2024, doi: 10.21275/SR24302013533.
    [46] Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P. Simoncelli, "Image quality assessment: from error visibility to structural similarity," IEEE transactions on image processing, vol. 13, no. 4, pp. 600-612, 2004, doi: 10.1109/TIP.2003.819861.
    [47] S. Samiee and K. Roth, "The influence of global marketing standardization on performance," Journal of marketing, vol. 56, no. 2, pp. 1-17, 1992, doi: 10.1177/00222429920560020.
    [48] A. Odena, V. Dumoulin, and C. Olah, "Deconvolution and checkerboard artifacts," Distill, vol. 1, no. 10, p. e3, 2016, doi: doi.org/10.23915/distill.
    [49] G. Baudat, D. Lavanchy, and G. Müller, "High-speed AI image space wavefront sensing using embedded computing: achieving 1000 frames per second," in Photonic Instrumentation Engineering XII, 2025, vol. 13373: SPIE, pp. 120-134, doi: 10.1117/12.3043339.
    [50] S. Chavhan, I. S. Deepika, D. Gupta, and J. J. Rodrigues, "Energy-Efficient-Enabled Edge-AI-IoT Integrated Traffic Incident Analysis and Avoidance of Secondary Incidents," IEEE Internet of Things Journal, 2025, doi: 10.1109/JIOT.2025.3555408.
    [51] S. B. Ghantasala and G. Singh, "Hybrid Machine Learning and Finite Element Modeling for Accurate Prediction of Sintering-Induced Deformation in Material Extrusion Additive Manufacturing," Acta Materialia, p. 121225, 2025, doi: 10.1016/j.actamat.2025.121225.
    [52] J. Sukhnandan and G. A. Drosopoulos, "A machine learning approach used to predict the peak displacement, base shear and fundamental frequency of multi-storey steel structures under seismic excitation," Structures, vol. 73, p. 108367, 2025, doi: 10.1016/j.istruc.2025.108367.

    QR CODE
    :::